A common industrial process for producing an amide compound involves Beckmann rearrangement of a corresponding oxime compound. For example, ε-caprolactam which is industrially useful is produced by Beckmann rearrangement of cyclohexanone oxime. Rearrangement catalysts used are generally concentrated sulfuric acid and oleum. Since these strong acids must be used in the stoichiometric amounts or more, they form a large amount of ammonium sulfate as a byproduct during neutralization. Although laurolactam, which is a starting material for Nylon 12, is also produced in a similar manner, the process is more complex because cyclododecanone oxime as an intermediate product has a high melting point. In producing ε-caprolactam, both cyclohexanone oxime and e-caprolactam have a relatively lower melting point, so that oxime formation or rearrangement can be conducted in a solvent-free system, but production of laurolactam requires a reaction solvent. This reaction solvent must be able to substantially dissolve cyclododecanone oxime and be inert to concentrated sulfuric acid or oleum, and therefore the selection of the solvent is considerably restricted.
Only two processes are known for industrially producing laurolactam from cyclododecanone and an aqueous solution of hydroxylamine. One is a process commercially developed by Degussa Company. This method is as follows. Cyclododecanone is converted into an oxime using isopropylcyclohexane as a solvent, and after separating layers, a resulting solution of cyclododecanone oxime in isopropylcyclohexane is slowly added to concentrated sulfuric acid at a low temperature to prepare a solution of a cyclododecanone oxime sulfate adduct in sulfuric acid. After separating and recovering isopropylcyclohexane, the residual solution of cyclododecanone oxime sulfate adduct in sulfuric acid is heated to initiate Beckmann rearrangement of the oxime. After the rearrangement reaction, water is added to the system to dilute sulfuric acid, and then, the laurolactam produced is extracted with an organic solvent. Here, the extraction solvent may be isopropylcyclohexane or cyclododecanone. The extraction solvent is recovered by distillation from the resulting extraction solution and then laurolactam in the residue is purified by distillation (see, Patent Reference No. 1).
This process does not generate ammonium sulfate as a byproduct in the rearrangement reaction step, but requires enormously large facilities and energy for treating a large amount of waste diluted sulfuric acid. Furthermore, since cyclododecanone reacts with concentrated sulfuric acid to form a byproduct, the oxime-forming reaction must be completed for eliminating residual cyclododecanone, but due to hydrophobicity of isopropylcyclohexane, a mass transfer rate is low in an oil-water interface, leading to a longer oxime-forming reaction. As a whole, the process involves many steps of separation, recovery and recycling of solvents and, therefore, requires considerably large equipment expenses and energy.
Another industrial process is that commercially developed by Ube Industries-EMS. This process utilizes the fact that cyclohexanone oxime and caprolactam are good solvents for cyclododecanone oxime and laurolactam, respectively (for example, see Patent Reference 2). Specifically, a mixture of cyclododecanone and cyclohexanone is blended with an aqueous solution of hydroxylamine to produce oximes. Cyclohexanone oxime produced has a low melting point and is a good solvent for cyclododecanone oxime, so that the reaction can be conducted at 100° C. or lower and at an ambient pressure. Furthermore, cyclohexanone oxime is adequately hydrophilic for the oxime-forming reaction to quickly proceed, and the mixture is transferred to the rearrangement step without residual cyclohexanone or cyclododecanone. A rearrangement catalyst used is concentrated sulfuric acid or oleum. Whereas laurolactam produced has a high melting point, it is highly soluble in caprolactam having a low melting point. Therefore, the reaction can carried out even at a temperature of 100° C. or lower. The resulting rearrangement reaction solution is neutralized with ammonia water and then extracted with an organic solvent. Caprolactam can be dissolved in water to some extent, but is extracted into an organic solvent due to salting-out effect of ammonium sulfate. Next, a large amount of water is added to the solution containing extracted laurolactam and caprolactam, and caprolactam is extracted into the aqueous phase. From the separated organic phase, the organic solvent is recovered and laurolactam is purified by distillation. The aqueous phase is concentrated and after removing impurities, caprolactam is purified.
This process is excellent in that laurolactam and caprolactam can be produced together. However, as a process for producing laurolactam, has the following problems; (1) separation and purification of caprolactam requires large amounts of equipment expenses, resulting in low investment efficiency and the process involves operations of low energy efficiency such as concentration of an aqueous solution of caprolactam; (2) there is a restriction to a production ratio of laurolactam/caprolactam; and (3) caprolactam is a low-value-added product in comparison with laurolactam and an use efficiency of hydroxylamine is low.
Recently, there have been intensely investigated rearrangement catalysts which do not require a large amount of sulfuric acid or oleum. As a system containing a strong acid, there have been reported a mixture of rhenium peroxide ammonium salt and trifluoromethane sulfonic acid (Non-Patent Reference 1), indium triflate (Non-Patent Reference 2) and ytterbium triflate (Non-Patent Reference 3). Known methods utilizing a system containing an acid and a dehydrating agent include a method of conducting rearrangement reaction using phosphorous pentoxide or a condensed phosphoric acid compound and a fluorine-free sulfonic anhydride or sulfocarboxylic anhydride in a N,N-disubstituted amide compound as a solvent (Patent References 3 and 4) and a method using a zeolite catalyst pre-treated with an aqueous acid-containing solution (Patent Reference 5). As methods that use no acids, there have been suggested a method of conducting rearrangement reaction in the presence of a rhenium compound and a nitrogen-containing heterocyclic compound (Patent References 6 and 7) and a method of using zinc oxide (Patent Reference 8). Patent Reference 9 has disclosed a method of reacting an oxime and a carboxylic acid in a carboxylic acid solvent using cyanuric chloride (trichlorotriazine) as a dehydrating agent, whereby producing an ester which is then subjected to rearrangement reaction. Patent Reference 10 has disclosed a method where an oxime hydrochloride is subjected to rearrangement using cyanuric chloride (trichlorotriazine) as an initiator.
Although some of these catalysts and manufacturing processes can provide a high rearrangement yield, these methods employ special catalysts and/or solvents, for which a recovering or recycling procedure is not disclosed, and these are, therefore, unestablished as an industrial process.
Patent Reference 11 has described Beckmann rearrangement of an oxime compound in a polar solvent, wherein a rearrangement catalyst used is an aromatic compound (1) containing, as aromatic-ring member, at least one carbon atom having a leaving group, (2) containing at least three aromatic-ring members which are either or both of heteroatoms or/and carbon atoms having an electron-withdrawing group, and (3) wherein, two of the heteroatoms and/or carbon atoms having an electron-withdrawing group are at the ortho- or para-position to the carbon atom having an electron-withdrawing group. A similar description can be found in Non-Patent Reference 4. Non-Patent Reference 5 discloses that a phosphoric acid salt having a heterocyclic structure similar to that in Patent Reference 11 is active for Beckmann rearrangement.
The catalyst disclosed in Patent Reference No. 11 is highly active for a rearrangement reaction of cyclododecanone oxime to provide laurolactam in a high yield, and is, therefore, suitable as a rearrangement reaction catalyst in producing laurolactam. However, a nitrile which is recommended as a solvent cannot be used for an oxime-forming reaction because it reacts with hydroxylamine to form an amidoxime. Furthermore, since it is susceptible to hydrolysis, the loss of the catalyst inevitably happens in the step of removing catalyst and the like. Since it is highly miscible with water, a process for dehydrating materials for rearrangement becomes complex. Therefore, for establishing a practically feasible industrial process, solvents and processes must be selected, in consideration of individual steps from starting materials to a final product including an oxime-forming step.
Patent Reference 1: Japanese examined patent publication No. S52-033118 (1977-033118).
Patent Reference 2: Japanese Laid-open patent publication No. H05-4964 (1993-4964).
Patent Reference 3: Japanese Laid-open patent publication No. 2001-302602.
Patent Reference 4: Japanese Laid-open patent publication No. 2001-302603.
Patent Reference 5: Japanese Laid-open patent publication No. 2001-072658.
Patent Reference 6: Japanese Laid-open patent publication No. H09-301951 (1997-301951).
Patent Reference 7: Japanese Laid-open patent publication No. H09-301952 (1997-301952).
Patent Reference 8: Japanese Laid-open patent publication No. 2001-019670.
Patent Reference 9: Japanese examined patent publication No. S46.23740 (1971-23740).
Patent Reference No. 10: Japanese examined patent publication No. S47-18114 (1972-18114).
Patent Reference No. 11: Japanese Laid-open patent publication No. 2006-219470.
Non-Patent Reference 1: K. Narasaka, et. al., Chemistry Letter, pp. 489-492 (1993).
Non-Patent Reference 2: J. S. Sandhu, et. al., Indian Journal of Chemistry, pp. 154-156 (2002).
Non-Patent Reference 3: J. S. Yadav, et. al., Journal of Chemical Research(S), pp. 236-238 (2002).
Non-Patent Reference 4: K. Ishihara, et. al., Journal of American Chemical Sociaty, pp. 11240-11241 (2005).
Non-Patent Reference 5: M. Zhu, et. al., Tetrahedron Letters, pp. 4861-4863 (2006).